<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
On 12/20/2011 07:47, Tigersharke . wrote:
<blockquote
cite="mid:CAHRv5yhTtkrRoicMJLJVK7vb+R5y+Ac6iQqc6xxncQTiS6aK9A@mail.gmail.com"
type="cite"><br>
<br>
<div class="gmail_quote">On Tue, Dec 20, 2011 at 6:32 AM, Lars
Engels <span dir="ltr"><<a moz-do-not-send="true"
href="mailto:lars.engels@0x20.net">lars.engels@0x20.net</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="HOEnZb">
<div class="h5">On Mon, Dec 19, 2011 at 01:39:26PM -0500,
Kris Moore wrote:<br>
><br>
> Testers,<br>
><br>
> Got a few files here I thought I'd make available for
testing to<br>
> evaluate putting them into 9.0.<br>
><br>
> This updated /etc/rc & /etc/rc.delay files change
how booting works, to<br>
> try and boot us faster to a GUI login. This happens
by delaying certain<br>
> services / networking from loading until after the
GUI starts. On my<br>
> slower netboot here, it shaved a 84 second boot down
to 45ish seconds,<br>
> almost a 50% reduction. I'll be curious to hear about
others results,<br>
> plus if any problems occur.<br>
><br>
> To test this out, first as root do:<br>
><br>
> # cp /etc/rc /etc/rc.orig<br>
><br>
> Then copy the attached rc & rc.delay to /etc/rc
and /etc/rc.delay<br>
> respectively.<br>
><br>
> After that you can reboot and see if it makes any
difference for you.<br>
><br>
> If you want to undo the patch, just do:<br>
><br>
> # cp /etc/rc.orig /etc/rc<br>
> # rm /etc/rc.delay<br>
><br>
> I'd be curious to hear about boot-times before and
after, I timed it<br>
> here from the boot-splash screen to the time the
login appeared for GDM.<br>
><br>
> Thanks!<br>
<br>
</div>
</div>
Kris,<br>
<br>
I'd vote against adding it to 9.0. PCBSD is already in the
last RC stage<br>
and adding this feature may be nice but can cause breakage for
many<br>
people.<br>
Personally I'd love to have a faster booting FreeBSD with
parallelized<br>
deterministic hardware sensing and parallelized rc scripts.<br>
But I always disliked the way Windows does it: Having a login
prompt and<br>
desktop very early but then wait for the network... IMHO
network<br>
connectivity should always be the first thing to start because
I don't<br>
want an offline machine. :)<br>
<br>
Cheers<br>
<span class="HOEnZb"><font color="#888888">Lars<br>
</font></span><br>
_______________________________________________<br>
Testing mailing list<br>
<a moz-do-not-send="true"
href="mailto:Testing@lists.pcbsd.org">Testing@lists.pcbsd.org</a><br>
<a moz-do-not-send="true"
href="http://lists.pcbsd.org/mailman/listinfo/testing"
target="_blank">http://lists.pcbsd.org/mailman/listinfo/testing</a><br>
<br>
</blockquote>
</div>
Hello<br>
<br>
I have not taken the time to test this. However, it seems like a
reasonable addition as a mechanism that would allow for the
configuration of the delayed loading of things. If inclusion would
delay v9.0 final, then make it available by other ways.. possibly
after more testing if other edge cases rear their ugly heads.<br>
<br>
Lars does make a valid point, if it were to be enabled by default
there are likely to be complaints that the configuration chosen
would not match the desires of any number of users.<br>
<br>
I definitely like the certainty of having network access right
away, but if I ran other things such as a bittorrent
server/client, or other various things that would delay my own
access to the desktop (or its immediate responsiveness), then
assuming there were no other detriments, I'd like to be able to
delay those types of things.<br>
<br>
Thanks for your hard work!<br>
<br>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">_______________________________________________
Testing mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Testing@lists.pcbsd.org">Testing@lists.pcbsd.org</a>
<a class="moz-txt-link-freetext" href="http://lists.pcbsd.org/mailman/listinfo/testing">http://lists.pcbsd.org/mailman/listinfo/testing</a>
</pre>
</blockquote>
<br>
Thanks for the feedback guys. I agree that this would be best for
9.1 to give it time to mature / fix weird issues :)<br>
<br>
I'll keep updating this thread with newer versions for testing
though, since this is of particular interest to me. FWIW, this
really doesn't delay networking at all. It just makes things run in
parallel, so that while X / GDM is loading, the networking is
simultaneously coming up, instead of wasting all the time waiting
for things to load in sequential order before we get to a GUI. <br>
<br>
<br>
<pre class="moz-signature" cols="72">--
Kris Moore
PC-BSD Software
iXsystems</pre>
</body>
</html>